Photonic hybrid quantum computing

Abstract

Photons are attractive carriers of quantum information: they travel at the speed of light, barely decohere, and can be handled without cryogenics. Yet, their interactions are so weak that scaling up photonic quantum computers remains challenging. This review explains how “hybrid” photonic encodings, where each logical qubit is an entangled combination of a single photon and a bosonic mode, such as a coherentstate qubit, can overcome this bottleneck. Three main photonic paradigms are surveyed: discretevariable schemes using single photons and linear optics, continuousvariable schemes based on squeezedlight cluster states, and bosonic codes such as the cat, binomial, and Gottesman-Kitaev-Preskill (GKP) encodings. Each offers either good robustness or good gate determinism but not both simultaneously. Hybrid qubits inherit the orthogonality and low loss sensitivity of single-photon encodings, together with the almost deterministic Bellstate measurements available for cat states. This enables “ballistic” measurement-based architectures in which most operations are passive, with little feedforward. By comparing several concrete fault-tolerant schemes, the review shows that hybrid architectures can tolerate photonloss rates at the percentage level per mode while using significantly fewer optical resource states than leading discrete or bosonic approaches. Crucially, the key ingredients—hybrid entangled pairs and qubit converters between singlephoton and coherentstate qubits—have already been demonstrated in tabletop experiments and may be distributed over long distances in hybrid quantumnetwork protocols. The review therefore positions hybrid photonic quantum computing not as a distant ideal but as a feasible route toward scalable, network-ready quantum processors built from light.